Dissertation > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Computer network > General issues > The application of computer network

Research and Application of personalized service based on Web log mining

Author LiuYuZuo
Tutor ChenYing
School Beijing Institute of Technology
Course Computer Science and Technology
Keywords Personalized service Recommended system Information Theory Clustering Web log mining Mixed Recommended
CLC TP393.09
Type Master's thesis
Year 2010
Downloads 140
Quotes 1
Download Dissertation

With the rapid development of Internet applications and the continued expansion of the explosive growth of the Web information \How to quickly and accurately provide users with personalized services become a challenging problem. Personalized services to improve the efficiency of Web utilization of resources and access to information, to meet the user's individual needs, and has important theoretical significance and practical value. Personalized Web services-based Web data mining research focus is the core content of the personalized service system. Research for Web mining and personalized service and the main challenge, this paper focuses on personalized service based on Web log mining and Web log mining data preprocessing, user interest model personalized recommendation system key technologies in-depth study of the findings and recommendation algorithm. This paper describes the main contents and results are as follows: 1. Analysis of Web log mining data preprocessing data sources constitute a data source format, a detailed description of the various stages of data preprocessing, gives a Web-based heuristic rules log data preprocessing algorithm. 2 user interest model method and found the technology to explore, and a brief introduction to their respective works, advantages and disadvantages, as well as applicable. Focus on the basic K-Means algorithm clustering technology, including process and the limitations of the algorithm. For the shortcomings of the algorithm, the improved density-based Adaptive K-Means algorithm to improve the quality of clustering, and its effectiveness is verified by experiment. Classification personalized recommendation technology to study and compare the rule-based, content-based, collaborative filtering, based on the characteristics of the information discussed mixed Recommended. Focus on several recommendation algorithm based on information theory analysis, the analysis of the inadequacies of the traditional algorithm. On this basis, the recommendation algorithm based on information theory and user clustering, in the slightly reduced the consideration of the performance of the system time, a significant improvement in the effectiveness and accuracy of the recommended. Design and personalized recommendation system prototype based on Web log mining. Improved data preprocessing algorithm, K-Means algorithm and recommendation algorithm based on a hybrid recommendation system framework, the focus used to solve the problem of unregistered users and new users to effectively recommend to improve the user registration Recommended accuracy.

Related Dissertations
More Dissertations